blob: 1a55df24a10ad913a814f8638a7c8b040960ace4 [file] [log] [blame]
/*
Copyright (c) 2011 Intel Corporation
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
* Neither the name of Intel Corporation nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#ifndef L
# define L(label) .L##label
#endif
#ifndef cfi_startproc
# define cfi_startproc .cfi_startproc
#endif
#ifndef cfi_endproc
# define cfi_endproc .cfi_endproc
#endif
#ifndef cfi_rel_offset
# define cfi_rel_offset(reg, off) .cfi_rel_offset reg, off
#endif
#ifndef cfi_restore
# define cfi_restore(reg) .cfi_restore reg
#endif
#ifndef cfi_adjust_cfa_offset
# define cfi_adjust_cfa_offset(off) .cfi_adjust_cfa_offset off
#endif
#ifndef ENTRY
# define ENTRY(name) \
.type name, @function; \
.globl name; \
.p2align 4; \
name: \
cfi_startproc
#endif
#ifndef END
# define END(name) \
cfi_endproc; \
.size name, .-name
#endif
#define CFI_PUSH(REG) \
cfi_adjust_cfa_offset (4); \
cfi_rel_offset (REG, 0)
#define CFI_POP(REG) \
cfi_adjust_cfa_offset (-4); \
cfi_restore (REG)
#define PUSH(REG) pushl REG; CFI_PUSH (REG)
#define POP(REG) popl REG; CFI_POP (REG)
#define PARMS 8
#define ENTRANCE PUSH(%edi);
#define RETURN POP(%edi); ret; CFI_PUSH(%edi);
#define STR1 PARMS
#define STR2 STR1+4
.text
ENTRY (wcsrchr)
ENTRANCE
mov STR1(%esp), %ecx
movd STR2(%esp), %xmm1
mov %ecx, %edi
punpckldq %xmm1, %xmm1
pxor %xmm2, %xmm2
punpckldq %xmm1, %xmm1
/* ECX has OFFSET. */
and $63, %ecx
cmp $48, %ecx
ja L(crosscache)
/* unaligned string. */
movdqu (%edi), %xmm0
pcmpeqd %xmm0, %xmm2
pcmpeqd %xmm1, %xmm0
/* Find where NULL is. */
pmovmskb %xmm2, %ecx
/* Check if there is a match. */
pmovmskb %xmm0, %eax
add $16, %edi
test %eax, %eax
jnz L(unaligned_match1)
test %ecx, %ecx
jnz L(return_null)
and $-16, %edi
PUSH (%esi)
xor %edx, %edx
jmp L(loop)
CFI_POP (%esi)
.p2align 4
L(unaligned_match1):
test %ecx, %ecx
jnz L(prolog_find_zero_1)
PUSH (%esi)
/* Save current match */
mov %eax, %edx
mov %edi, %esi
and $-16, %edi
jmp L(loop)
CFI_POP (%esi)
.p2align 4
L(crosscache):
/* Hancle unaligned string. */
and $15, %ecx
and $-16, %edi
pxor %xmm3, %xmm3
movdqa (%edi), %xmm0
pcmpeqd %xmm0, %xmm3
pcmpeqd %xmm1, %xmm0
/* Find where NULL is. */
pmovmskb %xmm3, %edx
/* Check if there is a match. */
pmovmskb %xmm0, %eax
/* Remove the leading bytes. */
shr %cl, %edx
shr %cl, %eax
add $16, %edi
test %eax, %eax
jnz L(unaligned_match)
test %edx, %edx
jnz L(return_null)
PUSH (%esi)
xor %edx, %edx
jmp L(loop)
CFI_POP (%esi)
.p2align 4
L(unaligned_match):
test %edx, %edx
jnz L(prolog_find_zero)
PUSH (%esi)
mov %eax, %edx
lea (%edi, %ecx), %esi
/* Loop start on aligned string. */
.p2align 4
L(loop):
movdqa (%edi), %xmm0
pcmpeqd %xmm0, %xmm2
add $16, %edi
pcmpeqd %xmm1, %xmm0
pmovmskb %xmm2, %ecx
pmovmskb %xmm0, %eax
or %eax, %ecx
jnz L(matches)
movdqa (%edi), %xmm3
pcmpeqd %xmm3, %xmm2
add $16, %edi
pcmpeqd %xmm1, %xmm3
pmovmskb %xmm2, %ecx
pmovmskb %xmm3, %eax
or %eax, %ecx
jnz L(matches)
movdqa (%edi), %xmm4
pcmpeqd %xmm4, %xmm2
add $16, %edi
pcmpeqd %xmm1, %xmm4
pmovmskb %xmm2, %ecx
pmovmskb %xmm4, %eax
or %eax, %ecx
jnz L(matches)
movdqa (%edi), %xmm5
pcmpeqd %xmm5, %xmm2
add $16, %edi
pcmpeqd %xmm1, %xmm5
pmovmskb %xmm2, %ecx
pmovmskb %xmm5, %eax
or %eax, %ecx
jz L(loop)
.p2align 4
L(matches):
test %eax, %eax
jnz L(match)
L(return_value):
test %edx, %edx
jz L(return_null_1)
mov %edx, %eax
mov %esi, %edi
POP (%esi)
test %ah, %ah
jnz L(match_third_or_fourth_wchar)
test $15 << 4, %al
jnz L(match_second_wchar)
lea -16(%edi), %eax
RETURN
CFI_PUSH (%esi)
.p2align 4
L(return_null_1):
POP (%esi)
xor %eax, %eax
RETURN
CFI_PUSH (%esi)
.p2align 4
L(match):
pmovmskb %xmm2, %ecx
test %ecx, %ecx
jnz L(find_zero)
/* save match info */
mov %eax, %edx
mov %edi, %esi
jmp L(loop)
.p2align 4
L(find_zero):
test %cl, %cl
jz L(find_zero_in_third_or_fourth_wchar)
test $15, %cl
jz L(find_zero_in_second_wchar)
and $1, %eax
jz L(return_value)
POP (%esi)
lea -16(%edi), %eax
RETURN
CFI_PUSH (%esi)
.p2align 4
L(find_zero_in_second_wchar):
and $(1 << 5) - 1, %eax
jz L(return_value)
POP (%esi)
test $15 << 4, %al
jnz L(match_second_wchar)
lea -16(%edi), %eax
RETURN
CFI_PUSH (%esi)
.p2align 4
L(find_zero_in_third_or_fourth_wchar):
test $15, %ch
jz L(find_zero_in_fourth_wchar)
and $(1 << 9) - 1, %eax
jz L(return_value)
POP (%esi)
test %ah, %ah
jnz L(match_third_wchar)
test $15 << 4, %al
jnz L(match_second_wchar)
lea -16(%edi), %eax
RETURN
CFI_PUSH (%esi)
.p2align 4
L(find_zero_in_fourth_wchar):
POP (%esi)
test %ah, %ah
jnz L(match_third_or_fourth_wchar)
test $15 << 4, %al
jnz L(match_second_wchar)
lea -16(%edi), %eax
RETURN
CFI_PUSH (%esi)
.p2align 4
L(match_second_wchar):
lea -12(%edi), %eax
RETURN
.p2align 4
L(match_third_or_fourth_wchar):
test $15 << 4, %ah
jnz L(match_fourth_wchar)
lea -8(%edi), %eax
RETURN
.p2align 4
L(match_third_wchar):
lea -8(%edi), %eax
RETURN
.p2align 4
L(match_fourth_wchar):
lea -4(%edi), %eax
RETURN
.p2align 4
L(return_null):
xor %eax, %eax
RETURN
.p2align 4
L(prolog_find_zero):
add %ecx, %edi
mov %edx, %ecx
L(prolog_find_zero_1):
test %cl, %cl
jz L(prolog_find_zero_in_third_or_fourth_wchar)
test $15, %cl
jz L(prolog_find_zero_in_second_wchar)
and $1, %eax
jz L(return_null)
lea -16(%edi), %eax
RETURN
.p2align 4
L(prolog_find_zero_in_second_wchar):
and $(1 << 5) - 1, %eax
jz L(return_null)
test $15 << 4, %al
jnz L(match_second_wchar)
lea -16(%edi), %eax
RETURN
.p2align 4
L(prolog_find_zero_in_third_or_fourth_wchar):
test $15, %ch
jz L(prolog_find_zero_in_fourth_wchar)
and $(1 << 9) - 1, %eax
jz L(return_null)
test %ah, %ah
jnz L(match_third_wchar)
test $15 << 4, %al
jnz L(match_second_wchar)
lea -16(%edi), %eax
RETURN
.p2align 4
L(prolog_find_zero_in_fourth_wchar):
test %ah, %ah
jnz L(match_third_or_fourth_wchar)
test $15 << 4, %al
jnz L(match_second_wchar)
lea -16(%edi), %eax
RETURN
END (wcsrchr)